Probiotics and testosterone in men: the whole chain, end to end
The entire "probiotics boost testosterone" claim originates in one mouse study, and the same strain then failed in men, in a trial run by the company that sells it. Here is every link in the chain.
The whole testosterone chain, link by link, built to be quoted rather than to rank: each link carries its own permanent anchor, and the caveat that qualifies the human trial sits in the same block as the trial, never below it.
Do probiotics increase testosterone?
No. The only strain tested head-on in men with change in total testosterone as the prospectively registered primary endpoint — L. reuteri ATCC PTA 6475 — failed at both 1×10⁹ and 1×10¹⁰ CFU/day in a trial sponsored by BioGaia AB, the strain's own manufacturer (PMID 38770015 (opens in a new tab); NCT04577625 (opens in a new tab)). The claim originates in a mouse study with 5–10 animals per group (PMID 24392159 (opens in a new tab)). The authors of the human trial call it a pilot and warn it may be underpowered, so it fails to support an effect rather than refuting a small one.
The testosterone chain, end to end
Link 1 — the mouse study. Outbred CD-1 male mice, 5–10 animals per group, L. reuteri ATCC PTA 6475 at 3.5×10⁵ organisms/mouse/day in drinking water from 8 weeks of age for 12–40 weeks. Serum testosterone rose (control diet p=0.0266; Western diet p=0.0037); testis weight p=0.0062 at 12 months; seminiferous tubule area p<0.0001 (PMID 24392159 (opens in a new tab), PLoS ONE 2014;9(1):e84877). Credit where it is due: testosterone was measured at a single timepoint (5 months), and the design included an E. coli K12 sham-organism control at the identical dose, which is a genuinely good feature. Zero human data.
Link 2 — the same strain, in men, with testosterone as the registered primary endpoint. 57 men aged 50–65 randomised, 12 weeks, 1×10⁹ and 1×10¹⁰ CFU/day versus placebo. Total testosterone (nM): placebo 17.88±6.16 → 17.66±8.34; low dose 18.74±6.16 → 18.90±4.96; high dose 16.29±6.84 → 15.97±5.91. Every arm, including placebo, also received vitamin D₃ 400 IU/day. The authors' own conclusion: "the present study does not support the hypothesis that a probiotic supplementation with L. reuteri ATCC PTA 6475 can increase testosterone levels in ageing men" (PMID 38770015 (opens in a new tab); NCT04577625 (opens in a new tab), registered 2020-09-30, before first enrolment 2020-11-05). SHBG, LH, FSH and DHEAS were all null in the same trial.
Link 3 — who ran it. The trial's lead sponsor is BioGaia AB, registered study class INDUSTRY, and co-author Jakob Axelsson is a BioGaia employee (NCT04577625 (opens in a new tab)). The party with maximal commercial incentive to find an effect tested its own strain in the target population and found none.
Link 4 — the "they under-dosed the humans" defence does not survive arithmetic. Per animal-day the human dose is four to five orders of magnitude higher than the mouse dose; per kilogram it is roughly one order of magnitude higher (mouse ≈1.2×10⁷/kg/day versus an 80 kg man at ≈1.25×10⁸/kg/day). Dose is not why the human trial failed. Duration remains genuinely unaddressed: the mice were dosed continuously from young adulthood for up to ten months of life, while the men got 12 weeks starting at about 57.
Link 5 — the two men-only trials that reported testosterone increases, and why neither counts. L. paracasei B21060 (Maretti 2017, n=41) is a self-described pilot using non-parametric within-group Wilcoxon comparisons with no multiplicity control, and the active arm is a four-ingredient synbiotic (arabinogalactan 1,243 mg + FOS 700 mg + L-glutamine 500 mg), so nothing is attributable to the strain (PMID 28245352 (opens in a new tab)). Weizmannia coagulans BC99 (n=72, PMID 39683384 (opens in a new tab)) assayed testosterone with a research-grade commercial kit and reported it only as a figure with no means, SDs or exact p-value; NCT06307821 (opens in a new tab) was registered after study start and testosterone is not a registered outcome at all; the registered sponsor is the manufacturer while the paper declares no conflicts; and both arms received 24 g/day milk protein.
Link 6 — the null results nobody quotes. Bacillus subtilis DE111 at 1×10⁹ CFU/day for 12 weeks in 25 Division I male baseball athletes: the testosterone rise was a main effect of time (offseason training) present in both arms (PMID 30049931 (opens in a new tab)). Inactivated B. coagulans for 2 weeks in 16 male soldiers: "no significant change (p's > 0.05) in any of the performance or blood variables" (PMID 31306390 (opens in a new tab)). DHT has been measured in men after a probiotic exactly once, and it was null — alongside testosterone, androstanediol glucuronide, androstenedione, DHEAS, SHBG and leptin, in 39 enrolled / 31 completed men aged 18–37 (PMID 17109580 (opens in a new tab)).
Link 7 — the largest human dataset points at women, explicitly not men. Gut microbial 5α-reductase type 1 gene abundance versus circulating DHT: women β 0.062 (95% CI 0.026–0.098), P=0.00066; no association in men. n=7,235 (4,338 men, 2,897 women), SCAPIS, metagenomics plus LC-MS/MS (PMID 41124778 (opens in a new tab)). Two of that paper's authors are inventors on a probiotic patent and have received probiotic research funding — and they still published the null in men.
Link 8 — the same product category is sold to lower testosterone in women. In PCOS, probiotics reduce total testosterone at low certainty: MD −0.40 ng/mL (−0.73 to −0.07), p=0.017, GRADE Low; synbiotics MD −0.13 (−0.18 to −0.09), p<0.001, GRADE Low; and probiotics increased SHBG by 25.40 nmol/L (12.50–38.30), p<0.001 (PMID 37829729 (opens in a new tab), Talebi S et al.; direction confirmed by PMID 31053754 (opens in a new tab)). Both marketing claims cannot be evidence-based.
Two things that have never been done at all. Aromatase activity in men has never been measured after a probiotic. And no probiotic has ever been tested in testosterone-deficient or hypogonadal men — one trial is now registered (84 overweight and obese men, four arms, serum androgens at baseline and end; IRCT20240313061277N1). That registered trial is the only scientifically legitimate open question left in the testosterone story.
The move to watch for. The genuinely positive male data in this literature is about sperm, in trials that measured no hormones at all (PMID 33842820 (opens in a new tab)). Sperm quality is not testosterone. Conflating them is the single most common move in this niche.
Where this outcome sits in the 22-row table
The grade is assigned on the hub’s 22-row table and argued above. The row is reproduced here unchanged, so the grade a reader quotes from this page is the same grade the table carries. What a grade measures is set out under how the grades work: it is a statement about the strength of the evidence for a named strain, never a statement that a product does anything.
| # | Outcome | Best grade at strain level | The strain, and the trial behind it | Buyable at the dose that was studied? |
|---|---|---|---|---|
| 13 | Testosterone and male hormones | Grade F for **Testosterone and male hormones** — nothing qualifies | — | Nothing qualifies |
The strains named above, and the labels in this audit that print them
Every row below is a line on a printed supplement panel, read off that product’s own teardown. A panel printing a designator is a fact about the panel. It is not evidence that the product was tested, that it delivers the amount that was studied, or that the strain does anything — several of the paragraphs above say in terms that it does not, and the dose is unknowable wherever the panel gives no per-organism quantity.
- L. reuteri ATCC PTA 6475
-
- BioGaia Gastrus® Probiotic Chewables (30 chewable tablets) Panel prints “L. reuteri ATCC PTA 6475” with no per-organism quantity — the amount is inside a blend total · this record carries a verification flag.
- BioGaia Osfortis® Probiotic Capsules (60 capsules) Panel prints “L. reuteri ATCC PTA 6475 (L. reuteri Osfortis®)” at 10 billion CFU.
- L. paracasei B21060
-
Not printed on any panel this audit has transcribed.
- Weizmannia coagulans BC99
-
Not printed on any panel this audit has transcribed.
- Bacillus subtilis DE111
-
- Goli Nutrition Pre, Pro, Postbiotic Gummies, 60 Count Panel prints “Bacillus subtilis DE111” at 10 mg.
- Legion Biome Probiotic Panel prints “Bacillus subtilis DE111” at 57.5 mg (5 billion CFU).
- OLLY Probiotic Gummy, Tropical Mango, 80 Count (40 servings) Panel prints “Bacillus subtilis DE111” at 1 billion live cultures per 2 gummies.
- Ortho Molecular Products Ortho Spore IG (SKU 476090) Panel prints “Bacillus subtilis DE111®” at 3 mg — 1 Billion CFU · this record carries a verification flag.
- Swolverine Probiotix Panel prints “Bacillus subtilis DE111®” with no per-organism quantity — the amount is inside a blend total · this record carries a verification flag.
- Thorne FloraSport 20B, 30 capsules (SKU SF818) Panel prints “Bacillus subtilis DE111” at 5 billion CFU · this record carries a verification flag.
- Zenwise Digestive Enzymes with Prebiotics + Probiotics (180 count) Panel prints “Bacillus subtilis DE111” with no per-organism quantity — the amount is inside a blend total · this record carries a verification flag.
- B. coagulans GBI-30, 6086
-
- Digestive Advantage Probiotics + Lactase Digestive Enzymes, 96 Capsules Panel prints “Bacillus coagulans GBI-30, 6086 (branded BC30)” at 500 million viable cells · this record carries a verification flag.
Read next
- Probiotics for the prostate: every trial, and what none of them shows No human randomised trial of any probiotic strain in men with benign prostatic hyperplasia is indexed in PubMed. What exists instead, in full.
- Probiotics and erectile function Not "no evidence exists" — the trials that exist are placebo-free, in the wrong populations, using unnamed strains, with unreported effect sizes.
- Probiotics, male fertility and semen parameters The only male-specific outcome in the corpus with repeated positive human data — and it is still Grade C, and still not sold in this market.
- The evidence hub All 22 outcomes in one table, the strain × outcome table, and every other outcome page.
- What to buy, and what not to The verdict, the five reasons a man arrives here, and the products ranked by how much their labels disclose. Two of the five reasons end in nothing.